CCF2019 Supporting Statement Part B Final

B. Collection of Information Employing Statistical Methods

1. Correctional Facility Universe and Respondent Selection

The Census of State and Federal Adult Correctional Facilities (CSFACF, hereafter referred to as the CCF) data collection is a complete enumeration of adult correctional facilities operated by state governments or the federal Bureau of Prisons (BOP). Facilities operated under contract for state and BOP authorities are included. Military and territorial prisons are excluded.

The 2019 CCF frame was developed from outreach to the 50 state Departments of Corrections (DOCs) and the BOP. Prior to this outreach, BJS compiled a roster of facilities likely to be in scope for the CCF. This was done by examining rosters of adult confinement and community-based correctional facilities (CBCF) used for BJS’s 2016 Survey of Prison Inmates (SPI; OMB No. 1121- 0152) and 2011-12 National Inmate Survey (NIS; OMB No. 1121-0311) collections, as well as searches of DOC websites. Additional preparatory work was completed by reviewing websites of the largest operators of private correctional facilities to identify any gaps in coverage from the DOC-reported information.

Using the compiled state-specific rosters, two lists were made – one for CBCFs and one for confinement facilities. At the same time roster development work was taking place for the CCF, the NIS was set to begin verification of a roster of confinement facilities that would be used as the universe for drawing a sample of facilities for the 2019/2020 collection cycle. In an effort to minimize burden on respondents, the two efforts ran simultaneously, with the CCF focusing on the community-based correctional facilities and the NIS on confinement facilities. Because NIS does not sample community-based correctional facilities, it would have been inappropriate for BJS to do outreach confirming these facilities in that outreach effort. For each state DOC, the two prepopulated lists of known facilities were sent separately. BOP was contacted through the NIS project and asked to confirm or update its list of confinement facilities. Work for both projects was conducted under an OMB generic clearance No. 1121-0339. Respondents were asked to confirm and edit, as needed, facility-specific information for their facilities. They were also asked to add any in-scope facility which did not appear on the prepopulated lists and respond to a subset of questions posed for the listed facilities. Requesting and including these facilities added by the DOCs is of particular importance to ensure complete coverage.

The resulting 2019 CCF frame file will be a current and complete universe list of confinement and community-based adult correctional facilities whose primary use is to house prisoners for state or federal correctional authorities. Any duplicates will be identified and removed by examining the file for identical or nearly identical addresses, phone numbers, or names of facilities.

While the CCF is a facility-based data collection, based on prior experience fielding this census, BJS anticipates that most state DOCs and the BOP will designate a single person as a respondent responsible for providing the data for all state DOC- or BOP-operated facilities holding prisoners under their authority. However, state respondents may not be able to provide data on facilities operated under contract. For privately-contracted facilities, there may be a single respondent at the corporate level (e.g., CoreCivic, Geo Group, Management and Training Corporation, or Dismas Charities) who can provide data for multiple facilities. In other cases, the BJS CCF data collection agent will need to contact the individual contracted facilities to obtain data. BJS estimates that there will be approximately 50 state DOC respondents, one BOP respondent, and an additional 400 respondents for contracted facilities (e.g., respondents at the corporation or facility level). Respondents will be asked to provide 34 data elements for each confinement facility and 13 data elements for each community-based correctional facility for which they are reporting.

BJS anticipates three possible response models for the 2019 CCF: (1) a single respondent is identified at the DOC, BOP, or private organization and that person provides data for each of the facility forms; (2) a point of contact (POC) is identified at the DOC, BOP, or private organization and that person sends out the forms to individuals at each facility to fill out, but the DOC, BOP, or private organization POC is ultimately responsible for submitting all of the data; and (3) respondents are identified by the DOC, BOP, or private organization for each facility, and the data collection agent must contact and obtain data from each individual. In response models (1) and (2), the respondents are known as central reporters.

As in previous administrations, the 2019 CCF will be a census rather than a sample survey. The reasons for conducting a census include:

• The CCF will be used as a sampling frame for other studies such as NIS and SPI.

• The CCF has not been collected in its entirety since 2005, and many changes to the universe have been made; therefore, a census, rather than a survey, gives a more complete picture of the changes.

• The small increase in effort to conduct a census over a sample will allow BJS to report on all adult state and federal correctional facilities across the 50 states, instead of being limited to national-level estimates for facility characteristics.

2. Procedures for Collecting Information

The CCF is designed as a multi-modal data collection, with the online data tool serving as the primary mode, and hardcopy form or data submission using various formats (e.g., ASCII, Excel, SAS, or SPSS) offered as alternatives. Initial contact for data collection will be with POCs at each state’s DOC. These POCs were identified during the facility roster phase of the CCF/NIS under the OMB generic clearance. This initial outreach will occur before data collection begins to notify our DOC POCs of the upcoming data collection. The national implementation of the CCF will commence in June 2019. Data collection, nonresponse follow-up, and data quality follow-up will last approximately seven months and will include a variety of mailings and telephone contacts. A brief description of all steps in the data collection protocol are provided below.

• Agency Head POC designation letter. In early June 2019, the Agency Head POC designation letter (Attachment 9) and POC designation form (Attachment 10) will be sent¹ to the Agency Heads at the DOCs. This letter will introduce the collection, provide a state-specific list of facilities identified during roster development, and provide a link to the CCF website that includes reference versions of the CCF questionnaires. The letter requests the designation of an agency POC who can provide facility-level data for each of their DOC-operated facilities. The Agency Head will be informed that BJS will collect data directly from correctional facilities contracted to hold inmates for their state. Ad hoc communications will occur with private organizations, as it may be necessary to coordinate with a private organization POC on which facilities he or she can report on behalf of, versus which facilities may need a facility-level POC.

• Pre-notification letter. In early June 2019, a pre-notification letter will be mailed to individual and central respondents for contracted facilities announcing the upcoming data collection (Attachment 11). Letters will include the facility or list of facilities for which the respondent will be asked to provide data.

• Invitation letter. In July, RTI will email one of two invitation letters (Attachments 12 and 13) to each designated respondent across the 50 states and the BOP. Letters will be tailored to the type of respondent (central or individual). The e-mail will stress the purpose and importance of the CCF and the need for participation. It will also notify the respondent of the collection due date, provide instructions (including the web address and the login credentials) for submitting the form(s) online, and provide an RTI contact if the respondent has any questions or comments.

• Nonresponse reminder letters and telephone contact. One week prior to the collection due date (mid-September 2019), RTI will e-mail the first reminder (Attachment 14) to designated respondents who have not yet submitted their form(s) as notification of the impending due date. Immediately following the collection due date, telephone calls will be made to non-respondents to remind them of the collection due date (see Attachment 15 for a sample script for these contacts). Two weeks after the collection due date, another reminder (Attachment 16) will be sent via USPS Priority Mail and e-mail. All reminders will contain information for completing the web instrument.

• Collection closeout letter/final reminder. In November 2019, two distinct closeout letters will be sent to notify individual and central non-respondents (Attachments 17 and 18) that the collection is coming to an end and that their response is needed within two weeks. Data collection will continue for approximately three more weeks to allow for receipt of any remaining questionnaires.

• Telephone and e-mail data quality follow-up. Within two weeks of receiving a completed form, data review will begin. During review, if data discrepancies or missing data values are discovered, BJS’s data collection agent will follow up with respondents via telephone or e-mail to clarify responses or obtain missing information (see Attachment 19 for a sample script for these contacts).

• Thank you letter. Within two weeks of the completion of data quality follow-up, a thank you letter (Attachment 20) will be mailed to respondents.

The table below provides an overview of data collection contacts with their corresponding attachment.

2019 CCF Contacts

Contact description		Attachment
Agency Head POC designation letter		9
Pre-notification letter		11
Invitation letter		12, 13
1st reminder e-mail		14
Start telephone nonresponse follow-up		15
2nd reminder letter and e-mail		16
Closeout/final reminder letters to individual and central respondents		17, 18
Telephone and e-mail data quality follow-up		19
Thank you letter		20

Data Entry

Respondents completing the web instrument will enter their responses directly into the online instrument. For those respondents returning a hardcopy form via mail or fax, RTI staff will enter the data elements into the web instrument form. To confirm that editing rules are being followed and that data elements are being keyed accurately, RTI will review frequencies for the entered data after the first 10 percent of cases are received. Any anomalies, inconsistencies, or unexpected values will be investigated and resolved.

Throughout the remainder of the data collection period, RTI and BJS will conduct regular data frequency reviews to evaluate the quality and completeness of data captured in both the web and hardcopy modes. Interim data files will be made available to BJS via an SFTP site when response rates reach 50%, 75%, and 90%.

Data Editing

RTI will attempt to reconcile missing or erroneous data through automated and manual edits of each form within two weeks of completion. In collaboration with BJS, RTI will develop a set of edits that will use other data provided by the respondent on the questionnaire to confirm acceptable responses or identify possible errors due to missing or inconsistent data elements. For example, if an initial screening question was left blank, but the follow-up questions were completed, a manual edit would be made to indicate the intended positive response to the screening question. Through this process, RTI can quickly identify which hardcopy questionnaires require follow-up and indicate the items that need clarification or retrieval from the respondent.

Data Retrieval

An RTI project liaison will attempt to verify or collect correct information with respondents when errors found due to missing or inconsistent data elements cannot be resolved through data editing. Throughout the data retrieval process, RTI will document the questions needing retrieval (e.g., missing or inconsistent data elements) then request clarification on the provided information, obtain values for missing data elements, and discuss any other issues related to the respondent’s submission.

3. Methods to Maximize Response Rates

BJS and RTI will undertake various activities to ensure that high response rates are achieved for the 2019 CCF. Based on previous iterations of the CCF, BJS anticipates a 100% unit response rate for state DOC- and BOP-operated facilities and an 80% response rate from contracted facilities.

Questionnaire development

The 2005 CCF questionnaire was reviewed to ensure the collection of the most pertinent information and remove any questions that had high non-response rates to reduce burden. RTI reviewed item non-response and data quality from the 2005 CCF. RTI also conducted a cognitive test with nine respondents of an initially-revised version of the CCF. The instrument was refined based upon the findings from the cognitive test and BJS’s subject matter expertise. BJS also conducted scoping interviews between September 2014 and May 2015 (OMB generic clearance No. 0607-0725) prior to RTI’s preparations. Finally, modifications to the instrument were made to include items addressing important issues, like veteran housing and inmate-to-security staff ratios, in the corrections field.

Because of the differences in operations and population served in confinement and community-based correctional facilities, the 2019 census is made up of two instruments. The community-based correctional facilities form is a subset of 13 questions from the confinement facility form. The revised questionnaires were reviewed on several occasions by BJS and RTI for ease of use, flow, and additional survey methodology best practices to ensure ease in administration.

Respondent identification and data collection support

In an effort to ensure maximum respondent participation, BJS and RTI will (1) conduct outreach to notify DOC POCs of the upcoming CCF and make contact as needed throughout the data collection period, (2) send data collection invitations which include login instructions for the web version of the CCF in both hardcopy through the USPS and in electronic format, (3) offer a response mode other than web or hardcopy forms if requested, and (4) provide technical support and other help with the CCF as needed.

Data collection monitoring

The procedures outlined in Section B.2 will ensure a thorough, multi-pronged approach to invite jurisdictions and facility operators to participate in the 2019 CCF. The RTI data collection manager will monitor unit response rates on a weekly basis and ensure each pending response is receiving the intended reminder communications. Additionally, agencies responsible for an especially large number of facilities will receive particular monitoring, since their lack of response could greatly impact national-level estimates. These states may receive ad hoc communications beyond the standard protocol in an effort to see if there are burden concerns or questions about reporting.

To promote 100% item completion by respondents, RTI will monitor item response rates as forms are submitted. Also, the web instrument will utilize prompts when a respondent overlooks an applicable question or provides a response inconsistent with other, related responses. RTI will use a survey management system linked to the web instrument to flag missing items and invalid responses. RTI will also flag missing items on hard copy submissions. The RTI data collection manager will oversee phone and e-mail outreach to respondents to clarify missing or invalid responses and to take corrective action. Any changes to responses obtained through this follow-up effort will be tracked and entered in the data collection database.

4. Testing of Procedures

The questions in the 2019 CCF instrument were predominately revised versions of the 2005 CCF questions. Based on scoping interviews conducted in 2014-2015, a review of the Association of State Correctional Administrators’ Performance Based Measurement System, and a data quality assessment of the 2005 CCF, revisions to the 2005 CJ-43 form were made by BJS, reviewed and revised by RTI, further refined by BJS, and then cognitively tested in late 2018.

Scoping interviews were conducted primarily with planning and research staff from 18 state DOCs from September 2014 to May 2015, by trained interviewers at the U.S. Census Bureau under an OMB generic clearance. The interviews were designed to explore respondents’ understanding of the topics, determine data availability, understand the vocabulary used by respondents, and explore questionnaire process issues. Interviewees gave feedback on the 2005 CJ-43 form and on the estimated burden associated with providing information for a single facility.

RTI also conducted a thorough data quality assessment of the 2005 CCF. The goal of this effort was to examine the extent of item nonresponse and reporting inconsistencies that arose due to (1) data processing errors; (2) ambiguity of instructions, question wording, or response options; and (3) other issues related to the response task items (e.g., respondent reporting behavior/reactions to question formats such as “mark all that apply” versus “select only one option”). Items with high nonresponse rates were flagged and considered for deletion. In May of 2018, RTI provided BJS the Data Quality Report of the 2005 CCF (Attachment 21).

The cognitive tests included nine participants and were conducted in October-December 2018 by three trained interviewers from RTI. Participants included planning and research staff from seven DOCs, and facility staff from one state-operated confinement facility and one privately-operated facility. Facilities for which participants provided data and feedback were a mix of community-based and confinement facilities, operated by either a DOC or private organizations contracted by a DOC. The cognitive test focused on (1) the clarity of the instructions and question wording; (2) respondents’ ability and willingness to apply the term definitions provided on the instrument when answering the questions; (3) the availability of data needed to provide accurate responses; and (4) the estimated burden associated with participation. In November of 2018, RTI provided BJS the Cognitive Testing Findings and Recommendations report (Attachment 22).

Prior to the national implementation of the 2019 CCF, BJS and RTI will conduct thorough testing of the web-based data collection administration system through systematic user testing, including testing skip patterns, ensuring seamless reporting of data, and testing back-end data checks on entered responses. BJS and RTI will also use respondent recruitment and support procedures similar to other successful BJS agency collections, such as MCI (OMB No. 1121-0094), and Annual Survey of Jails (OMB No. 1121-0094), which have been field-tested and successfully employed.

Additionally, RTI has developed and utilized web-based instruments that are substantially similar to the format and design for the 2019 CCF. The web-based administration procedures successfully employed in similar BJS collections, such as the Law Enforcement Management and Statistics Survey (OMB No. 1121-0240) and Annual Surveys of Probation and Parole (OMB No.1121-0064), will be followed to ensure the successful administration of the 2019 CCF.

5. Contacts for Statistical Aspects of the Data Collection

The Corrections Statistics Unit at BJS takes responsibility for the overall design and management of the CCF, including the development of the questionnaires, and the analysis and publication of the data. The BJS contact for the CCF is –

Laura Maruschak, Statistician

Corrections Unit

Bureau of Justice Statistics

810 7^th Street, NW

Washington, D.C. 20531

(202) 307-5986

[email protected]

The current data collection agent for the CCF is RTI. The contact information for the CCF project director at RTI is –

Chris Ellis, Senior Survey Director

RTI International

3040 E. Cornwallis Road

Research Triangle Park, NC 27709

(919) 541-6480

[email protected]

1 Unless otherwise noted, all mailings will be sent via USPS 1^st Class Mail.